1. Field of the Invention
The present invention relates to an ATM (Asynchronous Transfer Mode) communication system with a CLSF (Connection-Less Service Function).
2. Description of the Background Art
In order to provide the highly efficient and flexible communication services with respect to the increasing demands for the variety of communications such as the image communication and the high speed data communication, there is an eager expectation for the realization of the B-ISDN (Broadband-Integrated Service Digital Network), and the ATM exchange scheme is considered as a prospective scheme for actually realizing the B-ISDN.
The ATM exchange scheme is a scheme for realizing the communication service by loading data into a fixed length packet called cell regardless of the attributes of the data, and using this cell as a unit of exchange. The ITU (formerly CCITT) has formally determined this ATM exchange scheme as the next generation exchange scheme, and decided to use this ATM exchange scheme for realizing the B-ISDN. For this reason, it is highly likely that the demands for the next generation multi-media communication and broadband communication are going to be handled by constructing the public network or the local network based on the ATM exchange scheme.
In recent years, there is a movement for applying this ATM exchange scheme to the LAN (Local Area Network) such as the Ethernet. In this case, the LAN operated under the ATM exchange scheme will be referred as the ATM-LAN. Such an ATM-LAN is expected to have the advantages that the throughput of the LAN can be improved considerably, that it is suitable for the multi-media, and that it is adaptive to the public network.
Now, one of the features of the ATM communication scheme is that its high speed operation realized by the hardware switching of the ATM cells. That is, the ATM network is the connection-oriented (CO) network in which the virtual connection (VC) or the virtual path (VP) is set up end-to-end, and the packet called cell is delivered end-to-end by label multiplexing or label exchanging the VCs or VPs in terms of their identifiers (VCI or VPI).
The data to be delivered end-to-end is loaded in the payload section of the ATM cell, and the ATM cell is exchanged and transmitted up to the destination terminal by the hardware switching operation alone without the intervention of the software operation, where the hardware switching operation is carried out by the ATM switch according to the VPI/VCI (or the value of the other field such as PT in the ATM cell header) contained in the ATM cell header.
In contrast to this ATM communication scheme which is the connection-oriented communication scheme, the communication scheme used in the conventional data communication is the connection-less (CL) communication scheme in which the end-to-end connections are not necessarily set up, and the packet is transmitted to the destination terminal as the packet is sent out to the network by attaching the destination data as its part while some node in the network analyzes the destination data and carries out the routing processing. Namely, in the connection-less communication scheme, the data transmission is realized without the procedure for setting up the connections at the terminals. In such a case, The packet to be transmitted to the destination terminal in connection-less manner is called datagram and this data transmission is called the datagram transmission. Thus, in the connection-less communication scheme, the communication is realized in a form of the datagram transmission without the procedure for setting up the connections.
Almost all of the existing data terminals such as the workstation (WS) and the personal computer (PC) adopts this datagram transmission scheme because the datagram transmission scheme is supported by the LAN, and the software provided within the data terminals such as protocols TCP/IP and UDP/IP has been suitable for the datagram transmission.
In such existing terminals or terminals provided with the existing protocols, i.e., the terminal which generates the datagram and outputs it to the destination terminal/network through the ATM network, the datagram transmission scheme is used for the terminal to terminal communication. To this end, it is necessary for the terminal and the network to be modified to realize the function for adapting the terminal to the interface with respect to the ATM-LAN by replacing the usual LAN board with the ATM board such as the Ethernet board or by using the terminal adaptor (TA), the function for loading the datagram into the ATM cell somehow at the terminal, and the function to deliver the datagram to the destination terminal indicated by the destination address at the network. Here, the terminal include the gate-way between the existing LAN and the ATM network.
To realize these functions, the datagram delivery scheme using the CLSF has been used conventionally. In this datagram delivery scheme, the CLSF processing unit is provided within the ATM network, and all the datagrams are collected there once. In other words, the CLSF processing unit is connected with all the datagram terminals by PVC (Semi-Permanent VC) (or VC, VP, PVC, or PVP), and the terminal wishing to transmit the datagrams assembles the ATM cells for all the datagrams to be transmitted, and transmits the ATM cells to the VC directed toward the CLSF processing unit. The CLSF processing unit then reproduces the received datagrams, and selects the VC connected to the destination address by analyzing the destination address of the datagrams, and then re-assembles the ATM cells for the datagrams and transmits the ATM cells to the selected VC. In a case the VC connected to the destination address cannot be found while there are other CLSF processing units within the network, the CLSF processing unit transmits the re-assembled ATM cells to the next stage CLSF processing unit which is expected to contain the terminal with the destination address or which is determined by the routing rule in advance.
Here, it is not absolutely necessary for the CLSF processing unit to analyze the destination address after reproducing the datagrams, and transmit the ATM cells after re-assembling the ATM cells. For instance, in a case where the destination address is contained in the to cell among the ATM cells for the datagrams, the destination address of the first cell alone can be analyzed and then transmitted to the destination terminal, and then the subsequent cells of the ATM cells for the datagrams can be sequentially transmitted to that destination terminal.
However, in this datagram delivery scheme using CLSF, all the datagrams originating within the network are always going to be transmitted via the CLSF processing unit, so that the CLSF processing unit is required to have a higher throughput as the number of datagrams to be transmitted increases and as the number of terminals within the network increases. Consequently, the CLSF processing unit is required to have a very high throughput and the flexibly expandable.
Another scheme for transmitting the datagrams to the destination terminal is to set up an ATM connection, such as a VC, to the destination address, and the ATM cells for the datagrams are delivered through this VC. However, in this scheme, there is a serious problem concerning the selection of the destination terminal with respect to which the VC is to be set up. Namely, there are enormously many terminals to which the datagrams can possibly be transmitted in practice, and in addition the generation of the datagrams is more bursty compared with the speech data, etc., so that to set up the enormously many connections is going to be a considerable waste of the network resource.
Furthermore, in a case of realizing the connection-less communication in the conventional ATM network, the ATM connection is always terminated at the CLSF processing unit, and the protocol processing for the upper layers above the AAL layer such as the protocol for the connection-less service called CLNAP (Connection-Less Network Access Protocol) is carried out. In other words, even in a case of the datagram transmission between quite nearby terminals, the ATM connection is going to be terminated once at the CLSF processing unit. Also, in a case of using a datagram transmission between very distant terminals, it becomes necessary to pass through a plurality of CLSF processing units, each of which carries out the protocol processing above the AAL layer.
In general, the protocol processing above the AAL layer, such as the CLNAP, is realized by software processing so that the processing speed is slow compared with the processing below the AAL layer which is usually carried out by the hardware processing. Also, it is necessary for the CLSF processing unit to carry out the analysis of the address data such as the network layer address data in the datagrams, for not just the transmissions to the terminals of the network supported by that CLSF processing unit itself but also for the transmissions to the terminals of the network supported by the other CLSF processing units as well. This concentrates the datagram transmission processing load on the CLSF processing unit. For these reasons, it has been difficult to realize high speed communication in connection-less communication (datagram delivery) among terminals of a conventional ATM communication system.
On the other hand, in making the inter-LAN connection, i.e., inter-networking among the LANs, in the conventional LAN environment, a router has been required to be provided between each adjacent LAN. The main function of this router is the routing processing for the datagram transmission over the LANs, by processing up to the third layer (network layer) in the OSI (Open Systems Interconnection) protocol layer stack. Namely, for the datagram to be transmitted over two LANs, the datagram must be brought up to the third layer by the router to analyze the destination network layer address there, and then delivered to the destination LAN according to the result of this analysis. The function of this router also realized by the so called "gate-way" in the context of the computer communication, but the "gate-way" is formally defined as that which carries out the processing up to the seventh layer, so that the element for realizing this function will be called router in the following.
There is also an element called a "bridge" which has a similar function to the router in realizing the inter-LAN connection. In this bridge, in contrast to the router which determines the destination LAN by analyzing the destination network layer address, the destination LAN is determined by analyzing the data link layer address (MAC address). Namely, the bridge realizes the inter-LAN connection by analyzing the destination MAC address of the datagram and passing the datagram through to another LAN when the obtained MAC address is not destined within its own LAN.
Furthermore, there is also a similar element called "brouter" which functions as the router for the predetermined network layer protocol and as the bridge for all the other protocols.
These router, bridge, and brouter have been usually realized by the workstation (WS). Namely, the CPU provided within the WS carries out the address analysis and realized the functions of the router, bridge, and brouter by transmitting the datagram to the allocated physical port.
However, in a case of the ATM-LAN, these router, bridge, and brouter are going to terminate the connection at the third layer or the second layer 2 forcefully and the processing for the third layer and second layer after the termination is most likely handled by the software processing. For this reason, for the transmission over the LANs, the speed and the capacity of the communication can be considerably lowered compared with the communication within the LAN. Also, in a case of providing the router, bridge, brouter, etc., the VP/VC cannot be set up over the LANs because the layer processing above the ATM layer between the end points is carried out at the routers.
Thus, in the conventional ATM-LAN, the third layer (network layer) processing must be carried out at the physical boundary of the networks, because the physical network boundary is the boundary of the second OSI layer (data link layer), and therefore the router must be provided at the physical network boundary for this reason.
Moreover, the conventional routing protocol to be executed by the router cannot be executed correctly unless the router is located within the physical network to which it belongs.
As a consequence, the location of the router has been dictated by the physical configuration of the network conventionally, i.e., the topology of the network layer cannot be defined independently from the topology of the physical network. In addition, it has been impossible to locate the router belonging to a certain network outside of that certain network.